1

Encyclopedia of Life Sciences A21904

The ecological consequences of habitat fragmentation

Raphael K. Didham1,2,3

1School of Animal Biology, The University of Western Australia, 35 Stirling Highway,

Crawley WA 6009, Australia

2CSIRO Entomology, Centre for Environment and Life Sciences, Underwood Ave, Floreat

WA 6014, Australia

3School of Biological Sciences, University of Canterbury, Private Bag 4800, Christchurch,

New Zealand

*Correspondence: Professor Raphael K. Didham, raphael.didham@uwa.edu.au,

raphael.didham@csiro.au, Ph. +61 (0)8 9333 6762

2

Abstract

Habitat fragmentation is the process by which habitat loss results in the division of large,

continuous habitats into smaller, more isolated habitat fragments. Habitat fragmentation is

one of the most important processes contributing to population decline, biodiversity loss, and

alteration of community structure and ecosystem functioning in anthropogenically-modified

landscapes. Many thousands of individual scientific studies now show unequivocal evidence

for the impacts of patch area, edge effects, patch shape complexity, isolation, and landscape

matrix contrast on population and community dynamics in mosaic landscapes. However,

striking disparities in the results of individual studies across differing taxa and differing

ecosystems have raised considerable debate about the relative importance of the different

mechanisms underlying fragmentation effects, and even debate about the utility of the

‘habitat fragmentation’ concept in general. Resolution of this debate lies in clear

discrimination of the direct and indirect causal relationships among patch versus landscape

variables. The most important recent advances in our understanding of the ecological effects

of habitat fragmentation all stem from recognition of the strong context dependence of

ecosystem responses, including spatial context dependence at multiple scales, time lags in

population decline, trait dependence in species responses, and synergistic interactions

between habitat fragmentation and other components of global environmental change.

Keywords: connectivity, context dependence, edge effects, extinction debt, habitat area,

habitat fragmentation, habitat loss, isolation, landscape structure, matrix contrast, patch

shape, species traits, synergistic interactions.

3

Introduction

All ecosystems are heterogeneous at some spatial scale. One of the greatest advances in

theoretical ecology has been the realisation that spatial heterogeneity in environmental

conditions can fundamentally alter the outcome of species interactions and community

dynamics (Holt, 1984; Pickett & Cadenasso, 1995). Where patches of suitable versus

unsuitable habitat occur in a heterogeneous mosaic at spatial scales that are relevant to

resource acquisition, dispersal, reproduction, mortality, or other components of individual

fitness, then heterogeneity will impose selective pressures on organisms to adapt to spatial

patchiness in the landscape. In natural mosaic landscapes, organisms have been able to adapt

to the patchiness of available habitat over evolutionary time scales (e.g., through traits that

confer increased dispersal ability, or decreased resource specialisation), but this has not

typically been the case in landscapes modified by recent human influence. In

anthropogenically-modified landscapes, humans have been destroying habitat at rates that are

without precedent in the Earth’s evolutionary history (Skole & Tucker, 1993) – rates that far

exceed the capacity for most species to adapt and respond to the decreasing availability and

increasing patchiness of suitable habitat (Pimm et al., 1995; Myers & Knoll, 2001). ‘Habitat

fragmentation’, then, is an expression of the impact that rapid human alteration of the spatial

structure of the landscape has on ecological communities.

It is important to remember that habitat fragmentation represents just one subset of factors

that are more broadly encompassed under the impacts of global land-use change (Foley et al.,

2005). Land-use change refers to all components of change in the quantity and quality of land

cover types as habitat for organisms and productive land for humans. Although it would be

tempting to portray fragmentation of natural land cover types as the ‘quantity’ component of

land-use change (i.e. the amount and spatial arrangement of natural habitat remnants), and

4

production intensity on modified land cover types as the ‘quality’ component (i.e. the

increase or optimisation of inputs such as agrichemicals, fertilizer, water and energy to

maximise production outputs such as food, materials or human living space), the reality is

that there is no such simple dichotomy of effects in modified landscapes. Habitat

fragmentation and land-use intensification are integrally intertwined (Figure 1). Patterns of

habitat fragmentation have substantial effects on both internal habitat quality and external

crop production by humans, while at the same time patterns of land-use intensification

impose strong spatial structuring on communities in habitat remnants. Within the habitat

fragmentation literature, these interaction effects are typically dealt with by ignoring the

explicit mechanics of land management trade-offs that exist between maximisation of

production and minimisation of biodiversity loss, and simply treating land-use intensification

as one component of the external influences of the surrounding land-use matrix on processes

occurring within habitat remnants. This is not to say that these effects are considered

unimportant, but rather that emphasis is placed on the mechanistic details of patch dynamics

within production landscapes (a conservation-centred approach), as opposed to the

mechanistic details of how land-use productivity can be maximised while minimising

biodiversity loss (a sustainable production-centred approach).

Defining habitat fragmentation

The fragmentation concept – pattern or process?

Most authors define habitat fragmentation as the process by which habitat loss results in the

division of large, continuous habitats into a greater number of smaller patches of lower total

area, isolated from each other by a matrix of dissimilar habitats (modified from Wilcove et

al., 1986; Ranta et al., 1998; Franklin et al., 2002; Ewers & Didham, 2006a). This integrates

changes in the amount of remaining habitat as well as changes in the spatial arrangement of

5

that habitat under the same ‘umbrella’ term (Figure 2A). However, other authors have argued

that the term habitat fragmentation should be much more restricted in usage.

Some authors consider that fragmentation should be restricted to describing just one of five

precise ways in which individual units of habitat are broken up (perforation, dissection,

fragmentation, shrinkage or attrition; Forman, 1995; Collinge, 2009) because each process

has a different ecological effect. While this may be true in some instances, there are serious

limitations to any straight-forward separation of these small-scale processes, because

perforation, for example, might be a transient precursor to dissection of habitat, which is

itself a precursor to fragmentation, following which each subdivided piece of the dissected

habitat might be subject to shrinkage and attrition. Clearly, a small-scale reductionist

approach such as this is not helpful unless it is placed within a larger framework of spatial

and temporal changes at the landscape scale.

Other authors have recognised fragmentation as a landscape-scale process, but have argued

strongly for a dichotomy to be raised between habitat amount and habitat fragmentation as

two separate, and independent, effects on biodiversity in fragmented landscapes (Fahrig,

2003). Because all quantitative measures of spatial dispersion of habitat increase strongly

with decreasing habitat amount in the landscape, Fahrig (2003) and others have argued that

the ‘independent’ effects of habitat fragmentation can only be determined after first taking

into account the effects of habitat amount remaining in the landscape (Fahrig, 2003).

Effectively, these studies are arguing that fragmentation is not the key process operating in

fragmented landscapes – the key process is total habitat loss, irrespective of its dispersion in

space – and fragmentation should instead be restricted to describing the pattern of spatial

arrangement of remaining patches in the landscape following habitat loss (Figure 2B).

6

There is undoubtedly strong merit in explicitly considering the importance of habitat amount

in the landscape. Many, if not most, earlier studies of ‘fragmentation’ simply sampled a few

individual habitat patches and attributed all ecological effects to the spatial configuration of

the patches, without considering that some or all of these effects might actually have been

intercorrelated with other mechanisms related to the total amount of habitat remaining in the

landscape (Fahrig, 2003). On the other hand, the key conclusion of Fahrig (2003) and others,

that habitat fragmentation has a negligible effect on biodiversity after habitat amount is taken

into account represents a striking paradox when placed alongside the many thousands of

studies showing strong ecological effects of patch area, isolation, edge effects and other

factors. Reconciling this paradox requires explicit consideration of the direct and indirect

causal relationships between patch and landscape variables.

The underlying drivers of fragmentation - (pen)ultimate or proximate?

Consider a simple analogy for the debate about the ‘independent’ effects of habitat amount

versus habitat fragmentation per se: if land-use change goes hand in hand with human

population increase – that is, the two are completely intercorrelated with each other – can we

better understand the mechanistic effects of land-use change by concluding that human

population increase is the driver of change? The obvious answer is no. These drivers are

interlinked in a chain of ultimate versus proximate causality, and mechanistic understanding

only comes from identifying the proximate variables mediating the effect (Didham et al.,

2005). The same line of reasoning applies to the intercorrelation observed between

decreasing habitat amount and increasing habitat fragmentation in modified landscapes

(Figure 3A). Although habitat loss might ultimately (perhaps penultimately, if socioeconomic

human drivers are also considered to) be responsible for ecological effects in modified

7

landscapes, this does not improve mechanistic understanding without quantifying how these

effects are indirectly mediated by the altered spatial arrangement of habitat remnants (Figure

3B). Habitat loss acts via the change in habitat arrangement, not independently of it. The

direction of causality, and the temporal sequence of events, is clearly from habitat loss to the

resulting change in spatial arrangement and not the other way around, although even some

very influential publications have confused the direction of causality in some instances,

suggesting that “habitat loss and isolation... may result from, rather than [equate to],

fragmentation”; Forman, 1995, p.412). Of course, there can be very strong feedback effects

between these processes, as in cases where linear features such as roads or powerline

corridors initially create only a small amount of habitat loss, but result in increased edge

effects and isolation effects (as well as increased human access) that might feedback on

increased rates of habitat loss.

The relevance of fragmentation to conservation management – panchreston or paradigm?

Despite (or perhaps because of) the debate about how to properly define habitat

fragmentation, and the vast and ever-growing literature on the separate effects of patch area,

edge effects, patch shape complexity, isolation, and landscape matrix contrast, there is still a

definite need within conservation management for a single umbrella term to fully encompass

all of the patterns, processes and consequences of habitat change in modified landscapes.

This umbrella term could either be ‘habitat loss’ or ‘habitat fragmentation’, and it would not

matter which, as long as the underlying causal structure of variables is recognised (Figure

2b). Thus far habitat fragmentation has been the term that has resonated most widely across

the literature, in both a colloquial and a scientific sense. Far from habitat fragmentation being

a ‘panchreston’ – an archaic term that Lindenmayer & Fischer (2007) used to denote

fragmentation as a theory that is “made to fit all cases” and “used in such a variety of ways as

8

to become meaningless” – habitat fragmentation remains the most useful paradigm in

conservation management for describing the series of interlinked processes occurring in

modified landscapes and for enhancing communication among disciplines and the public

(Ewers & Didham, 2007b).

Patch dynamics in mosaic landscapes

Perhaps the greatest value of the seminal work by Fahrig (2003) is that she has cemented

fragmentation as the landscape-level phenomenon that it truly is, not as a patch-level

phenomenon. Much of what the study of habitat fragmentation is concerned with today is the

ecological consequences of land-use change for organisms living in networks of remnant

patches surrounded by a mosaic of modified or novel land use types. This was not always the

case, though. The historical roots of habitat fragmentation are embedded in the stochastic

spatial model of Island Biogeography Theory (IBT) (MacArthur & Wilson, 1967), which in

its strictest form considers just patch area and isolation, incorporates no external influence

beyond the probabilistic arrival of colonists across an inhospitable matrix and no internal

patch dynamics beyond probabilistic extinction rates, and is ‘neutral’ to species identities or

functional traits. Much has been made about the lack of relevance of habitat fragmentation to

landscape ecology because this underlying basis of IBT does not fit the complexity of

anthropogenically-modified landscapes, where strong external influences on patches are

paramount, and there is a blurring of the boundaries between what constitutes the ‘patch’ and

the ‘matrix’ from the niche perspective of an organism (Haila, 2002; Laurance, 2008).

However, the reality is that this argument is a ‘straw man’ in many ways – a situation that

never actually exists, but is set up simply to be falsified. There are no studies today that

assume a strict IBT framework, and all researchers recognise the external influence of the

surrounding matrix and the effects of variation in habitat quality and internal patch dynamics

9

on remnant populations. Perhaps in this sense, the criticisms of early IBT approaches have

been well heeded, and ‘mainstream’ habitat fragmentation studies are much more akin to

studies in other branches of landscape ecology than ever before.

Within the wider discipline, the five key indirect drivers of fragmentation effects are ever

more intensively studied, highlighting new aspects to the mechanistic basis for the effects of

altered spatial arrangement of habitat on population and community dynamics (for more

detailed discussion of these processes see reviews by Saunders et al., 1991; Andrén, 1997;

Didham, 1997; Tscharntke et al., 2002; Henle et al., 2004; Ries et al., 2004; Ewers &

Didham, 2006a; Kupfer et al., 2006; Banks-Leite & Ewers, 2009; Collinge, 2009).

Reduced patch area

Reduction in the total amount of habitat in the landscape inevitably leads to a strongly

skewed size distribution of remaining habitat patches, with many small patches and few large

patches (Ranta et al., 1998; Ewers & Didham, 2007a). The ecological effects of reduction in

patch area are more widely studied than for any other variable, except perhaps edge effects,

and this is largely attributable to the very long heritage of scientific interest in the underlying

causes of the species-area relationship (Lomolino, 2000). Species-area relationships have

been widely studied in the context of habitat fragmentation as well (Seabloom et al., 2002;

Ewers & Didham, 2006a), although this can be a contentious area of investigation

(Simberloff, 1992). Conceptually, any changes in species richness at the patch level are best

considered as an emergent property (or net outcome) of all the population-level and

community-level changes that combine to alter species occupancy in fragmented landscapes.

10

First, at the population level, the effect of altered spatial arrangement of habitat on the

distribution and abundance of individuals is the explicit theoretical domain of metapopulation

ecology (Hanski, 1998). Reduced patch area limits resource availability, reduces colonisation

rates, alters reproductive success, and imposes an intrinsic constraint on maximum population

size. At the extreme, this exposes populations to an increased risk of local extinction (Hanski

& Ovaskainen, 2000). The underlying mechanisms driving this relationship can be divided

into four categories – environmental stochasticity, demographic stochasticity, natural

catastrophes and reduced genetic diversity (Hanski & Gaggiotti, 2004) – but all four

processes have the potential to interact, creating what have been described as “extinction

vortices” (Gilpin & Soulé, 1986). These processes lie at the heart of population viability

analysis (Beissinger & McCullough, 2002), which comprises a set of analytical and

modelling techniques for predicting the probability of species extinction

Second, at the community level, there are typically large changes in species composition

associated with reduction in patch area (Saunders et al., 1991; Ewers & Didham, 2006a), as

different species have widely varying resource and area requirements, and differing dispersal

abilities. For instance, highly dispersive ground beetles are less affected by area reduction

than less dispersive groups, because increased dispersal rates can lead to the ‘rescue’ of small

populations that would otherwise have high extinction rates in small patches (de Vries et al.

1996). Because species with differing responses to patch area frequently interact with one

another (e.g. via predator-prey or competitive interactions), there is no simple way to predict

the community-level outcome of fragmentation from the sum of individual metapopulation

models. The incorporation of species interactions into the spatial context of metapopulations

is the theoretical domain of metacommunity ecology (Leibold et al., 2004; Gonzalez, 2009).

Metacommunity ecology has been responsible for the most significant re-invigoration of

11

spatial ecology in recent years, and is increasingly seen as the theoretical underpinning to the

study of habitat fragmentation (Gonzalez, 2009).

Finally, the net outcome of altered population and community dynamics is a characteristic

reduction in species richness in small habitat patches (Figure 4A) (Ewers & Didham, 2006a).

There are many potential explanations for a net positive species-area relationship such as this,

including strong influences of reduced habitat heterogeneity, reduced resource concentration,

increased disturbance, or altered colonisation-extinction dynamics (i.e. IBT) in small patches

(or more likely a combination of these factors operating on different species). Unlike oceanic

island systems, however, species-area relationships across habitat patches frequently show no

net relationship, or even a negative relationship (Cook et al., 2002; Ewers et al., 2007) when

there is an over-riding influence of external variables, such as context-dependence in the

effects of the surrounding landscape matrix on the ability of species to invade and occupy

small patches.

Increased edge effects

Habitat loss and reduction in patch area increase the proportion of habitat edge in the

landscape, and expose fragment interiors to external influence (Didham, 1997). Edge effects

describe the transition in abiotic and biotic variables that occurs across the boundary between

adjacent land-use types (Cadenasso et al., 2003). Edges are typically hotter, drier and windier

than the interior of patches, with a higher light intensity and modified plant composition and

habitat structure (Matlack, 1994; Williams-Linera et al., 1998; Chen et al., 1999; Harper et

al., 2005). These changes have pronounced effects on patterns of habitat use and the relative

abundance of animals at patch edges as well (Ries et al., 2004). Species richness typically

increases at the edge, sometimes dramatically so for many invertebrate taxa (Didham, 1997;

12

Ewers et al., 2007), and there can be substantial turnover in species composition (Ewers &

Didham, 2008). Ultimately, changes in both species richness and composition are a

composite of individual species responses, which are extremely varied both within and

between studies (Didham et al., 1998; Davies et al., 2000). As a generality, though, high

species richness at edges is typically a result of species influx from adjacent disturbed

habitats, whereas patch specialists typically decline at edges (Figure 4B).

A continued problem in interpreting these types of ecological edge effects is that most studies

have taken a ‘one-sided’ approach focusing on the internal dynamics of patch edges, without

considering the ‘two-sided’ nature of edge dynamics (Fonseca & Joner, 2007; Ewers &

Didham, 2008). This is particularly surprising as most drivers of edge effects originate

external to the patch. Moreover, because of some curious properties of factors such as

shading, wind turbulence, and spillover effects, the observed edge response for some

variables may actually occur well outside the structural vegetation edge of the patch

(Cadenasso et al., 1997), and therefore be missed entirely if sampling only occurs within the

patch. More generally speaking, the two-sided approach to edge studies places better

emphasis on the fact that edges are three-dimensional zones of transition between habitats

(Cadenasso et al., 2003), and have no absolute quantifiable dimensions unless comparisons

are made relative to both the adjacent patch and matrix interiors.

The other major recent advance in the study of edge effects is recognition that quantification

of edge impact requires explicit discrimination of two quite distinct components of edge

influence: edge extent (i.e. the distance over which a statistical difference in response can be

detected between the matrix and the patch) and edge magnitude (i.e. the degree of difference

in response between the patch interior and the matrix interior) (Harper et al., 2005; Ewers &

13

Didham, 2006b). For example, a particular edge response with a large extent but a small

magnitude of effect might not be considered to be as important, ecologically speaking, as an

edge response with a small extent, but a very large magnitude. The majority of previous

studies have focused on edge extent, rather than edge magnitude, even though it is the more

difficult variable to quantify (Ewers & Didham, 2006b). There are probably as many rules of

thumb about the distance of edge influence as there are edge studies (e.g., Skole & Tucker,

1993; Young & Mitchell, 1994; Chen et al., 1999). However, it is worth pointing out that the

reliability of these estimates is highly questionable, as most studies are limited in taking a

one-sided approach to the measurement of edge effects (Fonseca & Joner, 2007), most lack

the appropriate statistical rigour to determine edge extent reliably (Ewers & Didham, 2006b),

and edge extent is notoriously variable between different response variables and different

edge types (Laurance et al., 2002). Edge extent can vary from a few metres up to a kilometre

or more for different response variables (Ewers & Didham, 2008), and despite thousands of

individual studies of edge effects there has been relatively little progress toward

understanding the factors that determine edge extent (let alone appropriate integration of edge

extent and edge magnitude).

Altered patch shape

Shape complexity is a patch attribute that has been extremely poorly studied in comparison to

other components of habitat fragmentation (Ewers & Didham, 2006a). Fragments with

complex shapes have a much higher proportion of total fragment area that is edge, rather than

core habitat (Laurance & Yensen, 1991), and this has two important ecological consequences.

First, patches with higher shape complexity may have correspondingly higher patch

colonisation rates, and higher patch emigration rates, and this can cause greater variability in

population size and a decreased probability of population persistence (Hamazaki, 1996;

14

Collinge & Palmer, 2002; Cumming, 2002). Second, shape complexity accentuates the extent

to which edge effects permeate habitat patches (Collinge, 1996), reducing core area for patch

specialists. These effects are likely to be particularly severe for linear patch features, such as

strips of remnant vegetation along riparian zones. Surprisingly, these effects are also of

special concern for the world’s largest nature reserves. Recent studies suggest that a

characteristic feature of fragmented landscapes is that shape complexity increases

exponentially with increasing patch area (Ewers & Didham, 2007a), so that very large

patches contain far less core area than might otherwise be expected. Furthermore, the highly

convoluted nature of very large patches can result in the division of core habitat into multiple,

disjunct core areas that are separated by regions of edge-affected habitat (Ewers & Didham,

2007a). Population estimates based on a literature review of the density-area relationship

(Connor et al., 2000) showed that disjunct cores in large fragments can reduce population

size to one-fifth of that which could be supported if core habitat were continuous (Ewers &

Didham, 2007a). Taken together, the assumption is that these processes will lead to increased

species loss in patches with higher shape complexity (Figure 4C), but this has not been well

studied.

Increased patch isolation

Patch isolation – and its converse, patch connectivity – are not absolute quantities, rather

there are degrees of isolation in both time and space depending on the dispersal traits of the

species in question and their ability to cross the intervening matrix between patches. As

habitat loss increases in the landscape, both time since isolation and distance of isolation

increase in concert, and this can make it difficult to distinguish their effects (Saunders et al.,

1991). Immediately after isolation, a patch will typically contain far more species than will

be able to maintain viable populations in the long term, and species richness will decline

15

through time in a process called ‘species relaxation’ (Brooks et al., 1999). The time course

of relaxation depends greatly on the initial population size in the patch and the average

longevity of the organism, so can vary from months up to hundreds of years in the case of

long-lived trees (Vellend et al., 2006). Time of isolation can also be a significant factor

confounding comparisons of species diversity or composition between different patches if

some patches have been isolated for substantially longer than other patches.

The effects of time since isolation interact strongly with distance of isolation, as a narrow

spatial separation between patches may represent only a limited barrier to some organisms,

and allow populations to survive in the patch over the long term, when this might not

otherwise be the case in a more spatially isolated patch. However, gap or matrix crossing

abilities are highly variable between species. An extreme example of this was highlighted by

Bhattacharya et al. (2003) who found that two species of Bombus bumblebees would rarely

cross roads or railways despite the presence of suitable habitat that was within easy flying

range. Because some matrix habitats inhibit dispersal more than others (see Roland et al.,

2000; Ricketts, 2001) and because species differ in their ability to disperse through matrix

environments (Haddad & Baum, 1999; Collinge, 2000), the literature is full of seemingly

disparate results regarding the effects of spatial isolation on species and communities (Ewers

& Didham, 2006a). The most important conclusions from these studies are that spatial

isolation effects can depend as strongly on the structure of edge habitats as on the

permeability of the surrounding landscape, and the nature of species responses can depend as

much on behavioural predisposition to disperse as on physical ability to traverse large

distances.

16

Within conservation management, habitat corridors (either retained or restored) have been

widely promoted as a means of maintaining species diversity (Hilty et al., 2006) based on the

general principle that more connected patches have lower species loss rates than more

isolated patches (Figure 4D). Although there has been considerable debate over the potential

disadvantages of corridors, such as facilitation of the spread of disease, invasive species, fire

disturbance, and other threatening processes (Saunders & Hobbs, 1991), a wealth of studies

ranging from small-scale moss microecosystems (Gilbert et al., 1998) to large-scale forested

landscapes (Damschen et al., 2006) now show that corridors can increase population

abundance and species diversity in habitat patches, and even have a significant effect on the

outcome of species interactions and ecosystem functioning in fragmented landscapes

(Tewksbury et al., 2002; Levey et al., 2005).

Altered matrix structure

The ‘matrix’ is a term broadly used to describe the human-modified land-use types

surrounding remnant habitat patches, but variation in the structure and importance of the

matrix for different species defies any simple dichotomy in attributes between ‘patch’ and

‘matrix’. For many species, human-modified land-use types provide supplementary or

complementary resources that may compensate for limited resource availability in habitat

patches (Ries et al., 2004). Strictly speaking, then, there is no dividing line between what is a

‘patch’ and what is a ‘matrix’ in these cases, and some species may well perceive the whole

landscape as ‘habitat’ and there would be no ‘matrix’ per se. However, it is likely that even in

these cases habitat quality will still vary spatially in relation to the spatial distribution of

dominant land-use types. For example, Perfecto & Vandermeer (2002) demonstrated that ants

inhabiting forest patches in Mexican coffee plantations were actively foraging in the

surrounding coffee ‘matrix’ and that some species were even able to survive in this land-use

17

type in perpetuity. Furthermore, an increase in ‘matrix’ quality was associated with an

increase in the number of species and individuals that occurred there (Perfecto &

Vandermeer, 2002).

A growing body of evidence suggests that matrix quality is crucially important in

determining the abundance and composition of species within habitat patches (Gascon et al.,

1999; Cook et al., 2002). Matrix properties can affect the dispersal and movement of

individuals between patches (Gascon et al., 1999; Davies et al., 2001), and the degree of

structural contrast between patch and matrix determines the permeability of habitat edges to

propagule movement (Collinge & Palmer, 2002), which taken together can be the prime

determinants of colonisation-extinction dynamics (Brotons et al., 2003; Kupfer et al., 2006)

and species loss (Figure 4E). For these reasons the study of matrix structure is frequently seen

as the most important avenue for building a synthesis between patch and landscape

perspectives on habitat fragmentation. At the same time, modification of matrix quality in

order to facilitate dispersal, increase population size and increase the probability of

population persistence at both patch and landscape scales is seen as a viable practical strategy

for conservation management (Donald & Evans, 2006).

Confounding factors in the determination of species responses to habitat fragmentation

With the burgeoning fragmentation literature (Collinge, 2009) it is becoming ever more

difficult to interpret the huge number of case-specific ‘exceptions’ to the typical patterns of

response to fragmentation described above. In some cases the reason for this is that many

supposedly well-accepted generalities were actually founded on relatively weak evidence in

the first place. In other cases, the weight of evidence still supports perceived generalities, but

disparate results point to the need for a fundamental reconsideration of what we understand

18

about the mechanistic bases for fragmentation effects. In a recent review, Ewers & Didham

(2006a) provide a structured framework for interpreting the very large number of factors that

have been observed to confound straight-forward determination of the causal processes

underlying species responses to fragmentation. These factors include the degree of context-

dependence in the effects of individual components of fragmentation, the degree of temporal

dependence in ecological responses, the degree of trait-dependence of observed species

responses, and the degree to which fragmentation effects are dependent on, or interact with,

other components of global environmental change. Most of these areas remain entirely

unexplored, and represent the cutting edge of fragmentation research, and yet it is now clear

that understanding context-dependence, in all its many forms, will be crucial to understanding

the fragmentation process as a whole.

Interactions between multiple drivers of fragmentation

Recognition that “the matrix matters” (Ricketts, 2001) to our understanding of patch

dynamics is tacit acknowledgement that patch processes are dependent on, or interact with,

matrix context. Similar recognition of the interaction between matrix structure and the

magnitude of edge influence has pervaded the contentious ‘edge effects on nest predation’

literature (e.g., Tewksbury et al., 2006). Surprisingly, though, these examples of interactions

between multiple drivers of fragmentation have not been taken more broadly as being

indicative of the importance of spatial context-dependence among variables in general. For

example, Ewers et al. (2007) showed the striking importance of synergistic interactions

between patch area and edge effects on beetle communities in temperate forest remnants in

New Zealand. Despite thousands of individual studies investigating the effects of patch area

and thousands more studies separately investigating edge effects, only a handful of studies

have tested for interactions between the two variables (Ewers et al., 2007). In a similar vein,

19

the importance of patch shape complexity is likely to be entirely dependent on the magnitude

of edge effects, and similar arguments could be made for context-dependence among other

variables. No-one has yet tested for higher-order interactions among suites of variables, but

these almost certainly occur frequently, if not ubiquitously, as well.

Trait-dependence of species responses

A significant problem in interpreting the many seemingly contradictory species responses to

fragmentation is that the life history and biology of most species is so poorly known that it is

difficult to determine the mechanistic basis for population decline (or increase). In many

cases ecologists and conservation managers are also seeking more generality to their

conclusions about the effects of fragmentation than simply knowledge about how particular

species respond. Consequently, considerable effort has been expended drawing

generalisations about the traits of species that might confer susceptibility or resilience to

fragmentation (Henle et al., 2004). For example, a number of studies have found that habitat

specialists, species with large body size, species at higher trophic levels, and those with poor

dispersal abilities or a reliance on mutualist species are expected to go extinct first when

habitat area decreases (see reviews by Tscharntke et al., 2002; Henle et al., 2004; Ewers &

Didham, 2006a). In a similar vein, species at higher trophic levels (Zabel & Tscharntke,

1998) and species with ‘intermediate’ dispersal capabilities (Thomas, 2000) appear to be

most sensitive to patch isolation. Recognising these trait-dependencies can explain many of

the contrasting finds about the effects of patch area and isolation on species responses, and

the same is true of the other drivers of fragmentation effects also. Overall, across all drivers

of fragmentation effects, a very large number of traits have been proposed as being important

in determining species responses, although Henle et al. (2004) suggested that just six ‘traits’

(defined in the broadest sense) have sufficient empirical support to justify being considered

20

strong predictors of species’ sensitivity to fragmentation: population size, population

variability, competitive ability and sensitivity to disturbance, degree of habitat specialisation,

rarity, and biogeographic location.

Time lags in ecological response

Given the inherent difficulties in measuring temporal responses to human disturbance, the

issue of temporal context-dependence of fragmentation effects has been surprisingly widely

considered, if not always well studied directly. A range of transient dynamics has been

observed in habitat patches immediately following disturbance of the surrounding landscape,

with the most common of these being a short-term increase in population densities of mobile

taxa that ‘crowd’ into the remaining intact patches (Bierregaard et al., 1992; Collinge &

Forman, 1998). Ecological observations made during this time-frame can give a highly biased

impression of the ecological condition of the patch and the ecological impact of

fragmentation (Ewers & Didham, 2006a). Transient crowding effects may also impose a

strong destabilising effect on species interactions and community structure for resident

organisms within habitat patches, and might even be one of the factors that exacerbates the

rate or magnitude of community change in the long term.

Over the succeeding trajectory of species relaxation, populations decline and are lost at

varying rates depending on the spatial attributes of the remaining habitat and the traits of the

remaining species (Brooks et al., 1999; Ferraz et al., 2003), with rates being very rapid (on

the order of months to a few years) for short-lived invertebrate species in small habitat

patches (Collinge & Forman, 1998; Gonzalez, 2000). In some regions where habitat loss has

been exceptionally rapid and severe (such as the Atlantic Forest of Brazil, and many regions

of tropical South East Asia), there is considerable concern that long time-lags to extinction

21

for many taxa could mean that we are facing a very large ‘extinction debt’ (Tilman et al.,

1994; Brooks et al., 1999), that will inevitably lead to many hundreds, if not thousands, of

rare and endangered species disappearing over the coming century. Although at one time the

concept of extinction debts was largely theoretical (Tilman et al., 1994), there is a now a

growing body of empirical evidence supporting the strong temporal dependence of

community change following fragmentation (Ewers & Didham, 2006a; Kuussaari et al.,

2009). The same concept has also been extended to temporal dependence in a range of other

processes occurring concurrently in modified landscapes, including the concept of an

‘invasion debt’ (reflecting the number of species yet to invade a patch from incipient

populations in the landscape; Seabloom et al., 2006), and a ‘colonisation credit’ (reflecting

the number of species yet to colonize a patch during community re-assembly following

habitat restoration; Cristofoli et al., 2010), although these may eventually be seen as one in

the same concept (Jackson & Sax, 2010).

Synergistic interactions between fragmentation and other threatening processes

Finally, habitat fragmentation is not the only anthropogenically-driven process threatening

species in habitat patches, and in many cases it may not even be the most important factor.

Other components of land-use change, such as land-use intensification in the matrix for

example, can be the key determinant of patch processes (Tscharntke et al., 2005), and there

can be strong historical or environmental context dependence in the effects of other

components of global environmental change which blurs or confounds determination of

species responses to fragmentation (Didham et al., 2005; Didham et al., 2007). For instance,

in most, if not all, modified landscapes there is strong intercorrelation between loss of habitat,

invasion of non-native species and decline of native species, making it difficult to separate

whether it really is habitat loss or invasion that is the cause of decline (Didham et al., 2005).

22

Of even greater concern is the weight of emerging evidence showing that multiple drivers of

global change interact synergistically, rather than independently, such that the effects of one

driver exacerbate the effects of other drivers (Didham et al., 2007; Brook et al., 2008;

Tylianakis et al., 2008). Such synergies operate frequently, if not ubiquitously, between

habitat fragmentation and species invasion (Didham et al., 2007). For example, invasion of

the predatory beetle Coccinella septempunctata into cropland in the USA led to a three-fold

increase in predation pressure on a native aphid Bipersona sp. in natural grassland patches

within a crop matrix, compared to control sites in a more pristine grassland-dominated

landscape (Rand & Louda, 2006). Habitat fragmentation may also exacerbate future impacts

of climate change. For example, the geographical distribution of butterflies tracks climate

warming (Thomas et al., 2006) but in fragmented landscapes the movement of habitat

specialists is restricted by adverse conditions in the matrix, such that climate change impacts

are likely to be much more severe in heavily fragmented landscapes than in less modified

landscapes (Thomas et al., 2006). Numerous other examples of synergistic and antagonistic

interactions have also been recorded among all components of land-use change, atmospheric

CO2 increase, climate change, anthropogenic nitrogen deposition and species invasion

processes (Didham et al., 2007; Brook et al., 2008; Darling & Côté, 2008; Tylianakis et al.,

2008). Ecological understanding of the interactions among multiple drivers of global change

is still in its infancy and presents an important future challenge for conservation management

in fragmented landscapes (Didham et al., 2007; Tylianakis et al., 2008).

Conclusions

When considered in the broadest sense as an on-going process of human modification of the

amount, spatial arrangement and quality of semi-natural habitats remaining in the landscape,

‘habitat fragmentation’ continues to be the most widely-accepted and useful umbrella concept

23

for ecology, conservation and management in modified landscapes. The discipline of habitat

fragmentation is constantly developing, so that today most studies are either implicitly or

explicitly linking both patch and landscape perspectives on the ecological consequences of

land-use change for organisms living in networks of remnant patches surrounded by a mosaic

of modified or novel land use types. The most important recent advances in our

understanding of the ecological effects of habitat fragmentation all stem from recognition of

the strong context dependence of ecosystem responses, and the synergistic interactions

between habitat fragmentation and other components of global environmental change.

References

Andrén H (1997) Habitat fragmentation and changes in biodiversity. Ecological Bulletins 46,

171-181.

Banks-Leite C & Ewers RM (2009) Ecosystem Boundaries. In Encyclopedia of Life Sciences.

Chichester: John Wiley & Sons, Ltd.

Beissinger SR & McCullough DR (2002) Population Viability Analysis. Chicago, IL, USA:

University of Chicago Press.

Bhattacharya M, Primack RB & Gerwein J (2003) Are roads and railroads barriers to

bumblebee movement in a temperate suburban conservation area? Biological

Conservation 109, 37-45.

Bierregaard RO, Lovejoy TE, Kapos V, Santos AA & Hutchings RW (1992) The biological

dynamics of tropical forest fragments. A prospective comparison of fragments and

continuous forest. BioScience 42, 859-866.

Brook BW, Sodhi NS & Bradshaw CJA (2008) Synergies among extinction drivers under

global change. Trends in Ecology & Evolution 23, 453-460.

24

Brooks TM, Pimm SL & Oyugi JO (1999) Time lag between deforestation and bird

extinction in tropical forest fragments. Conservation Biology 13, 1140-1150.

Brotons L, Mönkkönen M & Martin J-L (2003) Are fragments islands? Landscape context

and density-area relationships in boreal forest birds. American Naturalist 162, 343-357.

Cadenasso ML, Pickett STA, Weathers KC & Jones CG (2003) A framework for a theory of

ecological boundaries. BioScience 53, 750-758.

Cadenasso ML, Traynor MM & Pickett STA (1997) Functional location of forest edges:

gradients of multiple physical factors. Canadian Journal of Forest Research 27, 774-782.

Chen J, Saunders SC, Crow TR, Naiman RJ, Brosofske KD, Mroz GD, Brookshire BL &

Franklin JF (1999) Microclimate in forest ecosystem and landscape ecology. BioScience

49, 288-297.

Collinge SK (1996) Ecological consequences of habitat fragmentation: implications for

landscape architecture and planning. Landscape and Urban Planning 36, 59-77.

Collinge SK (2000) Effects of grassland fragmentation on insect species loss, colonization,

and movement patterns. Ecology 81, 2211-2226.

Collinge SK (2009) Ecology of Fragmented Landscapes. Baltimore, MD, USA: The Johns

Hopkins University Press.

Collinge SK & Forman RTT (1998) A conceptual model of land conversion processes:

predictions and evidence from a microlandscape experiment with grassland insects. Oikos

82, 66-84.

Collinge SK & Palmer TM (2002) The influences of patch shape and boundary contrast on

insect response to fragmentation in California grasslands. Landscape Ecology 17, 647-

656.

Connor EF, Courtney AC & Yoder JM (2000) Individuals-area relationships: the relationship

between animal population density and area. Ecology 81, 734-748.

25

Cook WM, Lane KT, Foster BL & Holt RD (2002) Island theory, matrix effects and species

richness in habitat fragments. Ecology Letters 5, 619-623.

Cristofoli S, Piqueray J, Dufrêne M, Bizoux JP & Mahy G (2010) Colonization credit in

restored wet heathlands. Restoration Ecology in press.

Cumming GS (2002) Habitat shape, species invasions, and reserve design: insights from

simple models. Conservation Ecology 6, 3.

Damschen EI, Haddad NM, Orrock JL, Tewksbury JJ & Levey DJ (2006) Corridors increase

plant species richness at large scales. Science 313, 1284-1286.

Darling ES & Côté IM (2008) Quantifying the evidence for ecological synergies. Ecology

Letters 11, 1278-1286.

Davies KF, Margules CR & Lawrence JF (2000) Which traits of species predict population

declines in experimental forest fragments? Ecology 81, 1450-1461.

Davies KF, Melbourne BA & Margules CR (2001) Effects of within- and between-patch

processes on community dynamics in a fragmentation experiment. Ecology 82, 1830-

1846.

Didham RK (1997) An overview of invertebrate responses to forest fragmentation. In Forests

and Insects, pp. 303-320 [A Watt, NE Stork and M Hunter, editors]. London: Chapman

and Hall.

Didham RK, Hammond PM, Lawton JH, Eggleton P & Stork NE (1998) Beetle species

responses to tropical forest fragmentation. Ecological Monographs 68, 295-323.

Didham RK, Tylianakis JM, Gemmell NJ, Rand TA & Ewers RM (2007) Interactive effects

of habitat modification and species invasion on native species decline. Trends in Ecology

& Evolution 22, 489-496.

Didham RK, Tylianakis JM, Hutchison MA, Ewers RM & Gemmell NJ (2005) Are invasive

species the drivers of ecological change? Trends in Ecology & Evolution 20, 470-474.

26

Donald PF & Evans AD (2006) Habitat connectivity and matrix restoration: the wider

implications of agri-environment schemes. Journal of Applied Ecology 43, 209-218.

Ewers RM & Didham RK (2006a) Confounding factors in the detection of species responses

to habitat fragmentation. Biological Reviews 81, 117-142.

Ewers RM & Didham RK (2006b) Continuous response functions for quantifying the

strength of edge effects. Journal of Applied Ecology 43, 527-536.

Ewers RM & Didham RK (2007a) The effect of fragment shape and species' sensitivity to

habitat edges on animal population size. Conservation Biology 21, 926-936.

Ewers RM & Didham RK (2007b) Habitat fragmentation: panchreston or paradigm? Trends

in Ecology and Evolution 22, 511.

Ewers RM & Didham RK (2008) Pervasive impact of large-scale edge effects on a beetle

community. Proceedings of the National Academy of Sciences U.S.A. 105, 5426-5429.

Ewers RM, Thorpe S & Didham RK (2007) Synergistic interactions between edge and area

effects in a heavily fragmented landscape. Ecology 88, 96-106.

Fahrig L (2003) Effects of habitat fragmentation on biodiversity. Annual Review of Ecology

and Systematics 34, 487-515.

Ferraz G, Russell GJ, Stouffer PC, Bierregaard RO, Pimm SL & Lovejoy TE (2003) Rates of

species loss from Amazonian forest fragments. Proceedings of the National Academy of

Sciences of the United States of America 100, 14069-14073.

Foley JA, DeFries R, Asner GP, Barford C, Bonan G, Carpenter SR, Chapin FS, Coe MT,

Daily GC, Gibbs HK, Helkowski JH, Holloway T, Howard EA, Kucharik CJ, Monfreda

C, Patz JA, Prentice IC, Ramankutty N & Snyder PK (2005) Global consequences of land

use. Science 309, 570-574.

Fonseca CR & Joner F (2007) Two-sided edge effect studies and the restoration of

endangered ecosystems. Restoration Ecology 15, 613-619.

27

Forman RTT (1995) Land Mosaics: The Ecology of Landscapes and Regions. Cambridge,

UK: Cambridge University Press.

Franklin AB, Noon BR & George TL (2002) What is habitat fragmentation? In Effects of

Habitat Fragmentation on Birds in Western Landscapes: Contrasts with Paradigms from

the Eastern United States, pp. 20-29 [TL George and DS Dobkin, editors]: Cooper

Ornithological Society.

Gascon C, Lovejoy TE, Bierregaard RO, Malcolm JR, Stouffer PC, Vasconcelos HL,

Laurance WF, Zimmerman BL, Tocher M & Borges S (1999) Matrix effects and species

richness in tropical forest remnants. Biological Conservation 91, 223-229.

Gilbert F, Gonzalez A & Evans-Freke I (1998) Corridors maintain species richness in the

fragmented landscapes of a microecosystem. Proceedings of the Royal Society of London

Series B-Biological Sciences 265, 577-582.

Gilpin ME & Soulé ME (1986) Minimum viable populations: processes of species extinction.

In Conservation Biology. The Science of Scarcity and Diversity, pp. 19-34 [ME Soulé,

editor]. Massachusetts, USA: Sinauer associates, Inc.

Gonzalez A (2000) Community relaxation in fragmented landscapes: the relation between

species richness, area and age. Ecology Letters 3, 441-448.

Gonzalez A (2009) Metacommunities: spatial community ecology. In Encyclopedia of Life

Sciences. Chichester: John Wiley & Sons, Ltd.

Haddad NM & Baum KA (1999) An experimental test of corridor effects on butterfly

densities. Ecological Applications 9, 623-633.

Haila Y (2002) A conceptual genealogy of fragmentation research: from island biogeography

to landscape ecology. Ecological Applications 12, 321-334.

Hamazaki T (1996) Effects of patch shape on the number of organisms. Landscape Ecology

11, 299-306.

28

Hanski I (1998) Metapopulation dynamics. Nature 396, 41-49.

Hanski I & Gaggiotti OE (2004) Ecology, Genetics, and Evolution of Metapopulations. San

Diego, CA, USA: Elsevier.

Hanski I & Ovaskainen O (2000) The metapopulation capacity of a fragmented landscape.

Nature 404, 755-758.

Harper KA, MacDonald SE, Burton PJ, Chen J, Brosofske KD, Saunders SC, Euskirchen ES,

Roberts D, Jaiteh MS & Esseen PA (2005) Edge influence on forest structure and

composition in fragmented landscapes. Conservation Biology 19, 768-782.

Henle K, Davies KF, Kleyer M, Margules C & Settele J (2004) Predictors of species

sensitivity to fragmentation. Biodiversity and Conservation 13, 207-251.

Hilty JA, Lidicker WZ & Merenlender AM (2006) Corridor Ecology: the Science and

Practice of Linking Landscapes for Biodiversity Conservation. Washington, DC, USA:

Island Press.

Holt RD (1984) Spatial heterogeneity, indirect interactions, and the coexistence of prey

species. American Naturalist 124, 377-406.

Jackson ST & Sax DF (2010) Balancing biodiversity in a changing environment: extinction

debt, immigration credit and species turnover. Trends in Ecology & Evolution 25, 153-

160.

Kupfer JA, Malanson GP & Franklin SB (2006) Not seeing the ocean for the islands: the

mediating influence of matrix-based processes on forest fragmentation effects. Global

Ecology and Biogeography 15, 8-20.

Kuussaari M, Bommarco R, Heikkinen RK, Helm A, Krauss J, Lindborg R, Ockinger E,

Partel M, Pino J, Roda F, Stefanescu C, Teder T, Zobel M & Steffan-Dewenter I (2009)

Extinction debt: a challenge for biodiversity conservation. Trends in Ecology & Evolution

24, 564-571.

29

Laurance WF (2008) Theory meets reality: how habitat fragmentation research has

transcended island biogeographic theory. Biological Conservation 141, 1731-1744.

Laurance WF, Lovejoy TE, Vasconcelos HL, Bruna EM, Didham RK, Stouffer PC, Gascon

C, Bierregaard RO, Laurance SG & Sampaio E (2002) Ecosystem decay of Amazonian

forest fragments: A 22-year investigation. Conservation Biology 16, 605-618.

Laurance WF & Yensen E (1991) Predicting the impacts of edge effects in fragmented

habitats. Biological Conservation 55, 77-92.

Leibold MA, Holyoak M, Mouquet N, Amarasekare P, Chase JM, Hoopes MF, Holt RD,

Shurin JB, Law R, Tilman D, Loreau M & Gonzalez A (2004) The metacommunity

concept: a framework for multi-scale community ecology. Ecology Letters 7, 601-613.

Levey DJ, Bolker BM, Tewksbury JJ, Sargent S & Haddad NM (2005) Effects of landscape

corridors on seed dispersal by birds. Science 309, 146-148.

Lindenmayer DB & Fischer J (2007) Tackling the habitat fragmentation panchreston. Trends

in Ecology and Evolution 22, 127-132.

Lomolino MV (2000) Ecology's most general, yet protean pattern: the species-area

relationship. Journal of Biogeography 27, 17-26.

MacArthur RH & Wilson EO (1967) The theory of island biogeography. Princeton, N.J.:

Princeton University Press.

Matlack GR (1994) Vegetation dynamics of the forest edge - trends in space and successional

time. Journal of Ecology 82, 113-123.

Myers N & Knoll AH (2001) The biotic crisis and the future of evolution. Proceedings of the

National Academy of Sciences U.S.A. 98, 5389-5392.

Perfecto I & Vandermeer J (2002) Quality of agroecological matrix in a tropical montane

landscape: ants in coffee plantations in southern Mexico. Conservation Biology 16, 174-

182.

30

Pickett STA & Cadenasso ML (1995) Landscape ecology: spatial heterogeneity in ecological

systems. Science 269, 331-334.

Pimm SL, Russell GJ, Gittleman JL & Brooks TM (1995) The future of biodiversity. Science

269, 347-350.

Rand TA & Louda SA (2006) Spillover of agriculturally subsidized predators as a potential

threat to native insect herbivores in fragmented landscapes. Conservation Biology 20,

1720-1729.

Ranta P, Blom T, Niemelä J, Joensuu E & Siitonen M (1998) The fragmented Atlantic rain

forest of Brazil: size, shape and distribution of forest fragments. Biodiversity and

Conservation 7, 385-403.

Ricketts TH (2001) The matrix matters: effective isolation in fragmented landscapes. The

American Naturalist 158, 87-99.

Ries L, Fletcher RJJ, Battin J & Sisk TD (2004) Ecological responses to habitat edges:

mechanisms, models and variability explained. Annual Review of Ecology, Evolution and

Systematics 35, 491-522.

Roland J, Keyghobadi N & Fownes S (2000) Alpine Parnassius butterfly dispersal: effects of

landscape and population size. Ecology 81, 1642-1653.

Saunders DA & Hobbs RJ (1991) The role of corridors in conservation: what do we know

and where do we go? In Nature Conservation 2: The Role of Corridors, pp. 421-427 [DA

Saunders and RJ Hobbs, editors]. Australia: Surrey Beatty & Sons.

Saunders DA, Hobbs RJ & Margules CR (1991) Biological consequences of ecosystem

fragmentation: a review. Conservation Biology 5, 18-32.

Seabloom EW, Dobson AP & Stoms DM (2002) Extinction rates under nonrandom patterns

of habitat loss. Proceedings of the National Academy of Sciences U.S.A. 99, 11229-

11234.

31

Seabloom EW, Williams JW, Slayback D, Stoms DM, Viers JH & Dobson AP (2006) Human

impacts, plant invasion, and imperiled plant species in California. Ecological

Applications 16, 1338-1350.

Simberloff DS (1992) Do species-area curves predict extinction in fragmented forest? In

Tropical Deforestation and Species Extinction, pp. 75-89 [TC Whitmore and JA Sayer,

editors]. London: Chapman and Hall.

Skole D & Tucker CJ (1993) Tropical deforestation and habitat fragmentation in the

Amazon: Satellite data from 1978 to 1988. Science 260, 1904-1910.

Tewksbury J, Levey D, Haddad N, Sargent S, Orrock J, Weldon A, Danielson B, Brinkerhoff

J, Damschen E & Townsend P (2002) Corridors affect plants, animals, and their

interactions in fragmented landscapes. Proceedings of the National Academy of Sciences

U.S.A. 99, 12923-12926.

Tewksbury JJ, Garner L, Garner S, Lloyd JD, Saab V & Martin TE (2006) Tests of landscape

influence: Nest predation and brood parasitism in fragmented ecosystems. Ecology 87,

759-768.

Thomas CD (2000) Dispersal and extinction in fragmented landscapes. Proceedings of the

Royal Society of London Series B Biological Sciences 267, 139-145.

Thomas CD, Franco AMA & Hill JK (2006) Range contractions and extinction in the face of

climate change. Trends in Ecology and Evolution 21, 415-416.

Tilman D, May RM, Lehman CL & Nowak MA (1994) Habitat destruction and the extinction

debt. Nature 371, 65-66.

Tscharntke T, Klein AM, Kruess A, Steffan-Dewenter I & Thies C (2005) Landscape

perspectives on agricultural intensification and biodiversity - ecosystem service

management. Ecology Letters 8, 857-874.

32

Tscharntke T, Steffan-Dewenter I, Kruess A & Thies C (2002) Characteristics of insect

populations on habitat fragments: A mini review. Ecological Research 17, 229-239.

Tylianakis JM, Didham RK, Bascompte J & Wardle DA (2008) Global change and species

interactions in terrestrial ecosystems. Ecology Letters 11, 1351-1363.

Vellend M, Verheyen K, Jacquemyn H, Kolb A, Van Calster H, Peterken G & Hermy M

(2006) Extinction debt of forest plants persists for more than a century following habitat

fragmentation. Ecology 87, 542-548.

Wilcove DS, McLellan CH & Dobson AP (1986) Habitat fragmentation in the temperate

zone. In Conservation Biology. The Science of Scarcity and Diversity, pp. 237-256 [ME

Soulé, editor]. Massachusetts,USA: Sinauer associates, Inc.

Williams-Linera G, Domínguez-Gastelú V & García-Zurita ME (1998) Microenvironment

and floristics of different edges in a fragmented tropical rainforest. Conservation Biology

12, 1091-1102.

Young A & Mitchell N (1994) Microclimate and vegetation edge effects in a fragmented

podocarp-broadleaf forest in New Zealand. Biological Conservation 67, 63-72.

Zabel J & Tscharntke T (1998) Does fragmentation of Urtica habitats affect phytophagous

and predatory insects differentially? Oecologia 116, 419-425.

Further reading

Bierregaard RO Jr, Gascon C, Lovejoy TE and Mesquita R (2001) Lessons from Amazonia:

The Ecology and Conservation of a Fragmented Forest. Yale, CT, USA: Yale University

Press.

Gutzwiller KJ (2002) Applying Landscape Ecology in Biological Conservation. New York:

Springer.

Harris LD (1984) The Fragmented Forest. Chicago, IL, USA: University of Chicago Press.

33

Laurance WF and Bierregaard RO Jr (1997) Tropical Forest Remnants: Ecology,

Management and Conservation of Fragmented Communities. Chicago, IL, USA:

University of Chicago Press.

Lindenmayer D and Fischer J (2006) Habitat Fragmentation and Landscape Change: an

Ecological and Conservation Synthesis. Collingwood, Vic., Australia: CSIRO.

Rochelle JA, Lehmann LA and Wisniewski J (1999) Forest Fragmentation: Wildlife and

Management Implications. Boston, MA: Brill.

Schelhas J and Greenberg R (1996) Forest Patches in Tropical Landscapes. Washington,

DC, USA: Island Press.

Wiens JA and Moss MR (2005) Issues and Perspectives in Landscape Ecology. Cambridge,

UK: Cambridge University Press.

34

Figure captions

Figure 1. Land-use change in a fragmented tropical forest landscape in the highlands near

Sapa, Vietnam, showing the scale of habitat loss, the altered spatial arrangement of remaining

rainforest habitats, and the mosaic of human land-use intensification in the surrounding

landscape matrix. Photo reproduced with permission from iStock Photo

(http://www.istockphoto.com).

Figure 2. Defining habitat fragmentation. There is strong debate about whether the term

‘habitat fragmentation’ should be used to describe: (A) the entire spatio-temporal process by

which habitat loss leads to the subdivision of large, continuous habitats into a greater number

of smaller patches of lower total area, isolated from each other by a matrix of dissimilar

habitats; or (B) solely the differences that occur due to the differing pattern of spatial

arrangement of remaining habitat after the amount of habitat remaining in the landscape has

been taken into account. The satellite images in (A) show typical correlated changes in

habitat loss and habitat fragmentation in the same landscape east of Santa Cruz, Bolivia, over

three time intervals (images courtesy of NASA/Goddard Space Flight Center Scientific

Visualization Studio; http://visibleearth.nasa.gov). The satellite images in (B) show three

different landscapes in southern Mato Grosso, Brazil, with approximately the same amount of

habitat loss, but very differing spatial arrangement (images courtesy of Jacques Descloitres,

MODIS Land Rapid Response Team, NASA/GSFC, 29 May 2001;

http://visibleearth.nasa.gov).

Figure 3. A schematic representation of the problem of attributing causality to ‘habitat loss’

versus ‘habitat fragmentation per se’. (A) In modified landscapes, all measures of spatial

habitat configuration are strongly intercorrelated with the amount of remaining habitat,

35

making separation of ‘independent’ effects impossible. (B) The reason for the strong

intercorrelation is that the effects of habitat amount do not only operate directly and

separately from the effects of habitat fragmentation, they predominantly operate through

indirect pathways mediated by altered spatial configuration. For clarity, not all possible

indirect pathways are shown in (B).

Figure 4. Widely-held generalisations about community responses to habitat fragmentation.

Predictions of how species richness typically changes as the five main components of the

spatial context of habitat fragments are altered. Redrawn from Ewers & Didham (2006a).

36

Figure 1

37

Figure 2

Fragmentation as pattern

Fragmentation as process

(A)

(B)

38

Figure 3

39

Figure 4

Mag

nitu

de o

f sp

ecie

s lo

ss high

low

small interiorlarge edge low high connected isolated similar

Fragment area

Distance from edge

Shapecomplexity

Isolation Matrixcontrast

different

(A) (E)(C) (D) (B)